Regression Part-4

Question 1

What is the difference between the data collected in observational sampling and designed sampling experiments?

Answer 1

Analyst cannot control the variables under study in one while in another we can control the levels of variables to determine their effect on variables

Question 2

What is a Dummy Variable Trap?

Answer 2

When we have the same number of dummies as there are categories it creates a perfect collinearity problem

Question 3

What happens to the dummy we use as reference?

Answer 3

That dummy variable is omitted to avoid the DVT

Question 4

When is ANOVA used?

Answer 4

When we need to compare the means of 2 or more populations

Question 5

Which component describes the mean of the base category of the dummy variable?

Answer 5

the intercept

Question 6

What are the coefficients attached to the dummy variables called?

Answer 6

differential intercept coefficients and not the slope

Question 7

How can we circumvent DVT?

Answer 7

By removing intercept (check what other problems may arise as a result)

Question 8

What are main differences between intercept less and with intercept models?

Answer 8

1. DF = n-1 / DF = n-2; 2. sum of residuals may not be zero/always zero; 3. r2 can be negative so only use raw r2 value/ r2 cant be negative; 4. Use raw sum of squares/mean adjusted sum of squares

Question 9

What is an ANCOVA model?

Answer 9

Which contains both quantitative and qualitative variables. We can control the effects of covariates or control variables

Question 10

What is the principle of marginality?

Answer 10

If the interaction of X1 and X2 are statistically significant then we retain both X1 and X2 even if they are not statistically significant

Question 11

What is the similarity between concurrent and coincident regressions?

Answer 11

Both types mean that intercepts are the same

Question 12

Why do we add Dummy Variable D in the interaction term when comparing two regressions/two categories?

Answer 12

To differentiate between the slope coefficients

Question 13

Why do we add Dummy Variable D in the additive form term when comparing two regressions/two categories?

Answer 13

To distinguish between intercepts

Question 14

When comparing two regressions, when are they concurrent

Answer 14

When differential slope coefficient of dummy is statistically insignificant, so they have same intercept

Question 15

when comparing two regressions, when are they coincident

Answer 15

when both the differential slope coefficient of dummy and interaction is zero they have both slope and intercept same

Question 15

when comparing two regressions, when are they parallel

Answer 15

When differential slope coefficient of X variable is statistically insignificant but differential slope coefficient of dummy is statistically significant , so they have same slope

Question 16

What are the 4 components of a time series data

Answer 16

Trend, Seasonality, Cyclical, Random variation